RESEARCH Open Access

Quercetin prevents necroptosis ofoligodendrocytes by inhibitingmacrophages/microglia polarization to M1phenotype after spinal cord injury in ratsHong Fan1,2†, Hai-Bin Tang3†, Le-Qun Shan1, Shi-Chang Liu1, Da-Geng Huang1, Xun Chen4, Zhe Chen1,Ming Yang1, Xin-Hua Yin1, Hao Yang1* and Ding-Jun Hao1*

Abstract

Background: Oligodendrocytes (OLs) death after spinal cord injury (SCI) contributes to demyelination, even leadingto a permanent neurological deficit. Besides apoptosis, our previous study demonstrated that OLs underwentreceptor-interacting serine-threonine kinase 3(RIP3)/mixed lineage kinase domain-like protein (MLKL)-mediatednecroptosis. Considering that necroptosis is always accompanied with pro-inflammatory response and quercetinhas long been used as anti-inflammatory agent, in the present study we investigated whether quercetin couldinhibit necroptosis of OLs and suppress the M1 macrophages/microglia-mediated immune response after SCI aswell as the possible mechanism.

Methods: In this study, we applied quercetin, an important flavonoid component of various herbs, to treat ratswith SCI and rats injected with saline were employed as the control group. Locomotor functional recovery wasevaluated using Basso-Beattie-Bresnahan (BBB) scoring and rump-height Index (RHI) assay. In vivo, the necroptosis,apoptosis, and regeneration of OLs were detected by immunohistochemistry, 5′-bromo-2′-deoxyuridine (BrdU)incorporation. The loss of myelin and axons after SCI were evaluated by Luxol fast blue (LFB) staining,immunohistochemistry, and electron microscopic study. The polarization of macrophages/microglia after SCIand the underlying mechanisms were detected by quantitative reverse transcription-polymerase chain reaction(qRT-PCR) and immunohistochemistry. In vitro, the ATP and reactive oxygen species (ROS) level examination, propidiumiodide (PI) labeling, and Western blotting were used to analyze the necroptosis of cultured OLs, while the signalingpathways-mediated polarization of cultured macrophages/microglia was detected by qRT-PCR and Western blotting.

Results: We demonstrated that quercetin treatment improved functional recovery in rats after SCI. We then found thatquercetin significantly reduced necroptosis of OLs after SCI without influencing apoptosis and regeneration of OLs.Meanwhile, myelin loss and axon loss were also significantly reduced in quercetin-treated rats, as compared to SCI +saline control. Further, we revealed that quercetin could suppress macrophages/microglia polarized to M1 phenotypethrough inhibition of STAT1 and NF-κB pathway in vivo and in vitro, which contributes to the decreased necroptosis ofOLs.

Conclusions: Quercetin treatment alleviated necroptosis of OLs partially by inhibiting M1 macrophages/microgliapolarization after SCI. Our findings suggest that necroptosis of OLs may be a potential therapeutic target for clinical SCI.

Keywords: Spinal cord injury, Oligodendrocyte, Necroptosis, Macrophages/microglia, Quercetin

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

* Correspondence: yanghao.71_99@yahoo.com; haodingjun@126.com†Hong Fan and Hai-Bin Tang contributed equally to this work.1Shaanxi Spine Medicine Research Center, Translational Medicine Center,Department of Spine Surgery, Hong Hui Hospital, Xi’an Jiaotong University,555 You Yi Dong Road, Xi’an 710054, Shaanxi, ChinaFull list of author information is available at the end of the article

Fan et al. Journal of Neuroinflammation (2019) 16:206 https://doi.org/10.1186/s12974-019-1613-2

BackgroundThe secondary injury after spinal cord injury (SCI), moretightly related to severity of neurological deficit, includescell death, demyelination, axonal degeneration, and in-flammatory response [1–3]. Among these, oligodendro-cytes (OLs) death after SCI contributes to demyelinationof spared axons, even leading to a permanent neuro-logical deficit [4].Given that the survival of OLs is crucial for functional

recovery after SCI, in the present study we focused ondeath and regeneration of OLs. Apoptosis of OLs occursmainly in the first several hours after SCI, and the injuryis accompanied by chronic inflammatory demyelination[5–7], indicating other types of cell death accounted forthe delayed loss of OLs. It has been reported thatnecroptosis of OLs mediates axonal degeneration in ALS[8]. Our previous study demonstrated that some OLswere receptor-interacting serine-threonine kinase3(RIP3)/mixed lineage kinase domain-like protein(MLKL)-positive after SCI, indicating that necroptosisoccurred in OLs [9]. The chronic, detrimental, M1 mac-rophages/microglia-mediated inflammatory responsecontributed to necroptosis of astrocytes after SCI [10, 11].Likewise, the dominance of M1 sub-population might playa role in necroptosis of OLs. Thus, there is an urgent needfor new treatment strategies to prevent OLs death and toameliorate the M1 macrophages/microglia-mediated im-mune response after SCI.Quercetin, an important flavonoid component of

various herbs, has long been used as antioxidant andanti-inflammatory agent in traditional Chinese medi-cine [12–14]. It was confirmed to have neuroprotec-tive effects on SCI by reducing death of neurons,decreasing neural tissue damage, and inhibiting in-flammatory responses [15–18]. However, the effects ofquercetin on survival of OLs and polarization of mac-rophages/microglia after SCI still remain unclear.In this study, we investigated whether quercetin pro-

moted OLs survival and suppressed the detrimental M1-type immune response after SCI, and explored the pos-sible mechanism. We first showed that quercetin im-proved functional recovery after SCI. We thendemonstrated that quercetin significantly reducednecroptosis of OLs after SCI without affecting apoptosisand regeneration of OLs. Meanwhile, quercetin signifi-cantly reduced myelin and axon loss after SCI. We fur-ther revealed that quercetin suppressed macrophages/microglia polarized to M1 phenotype through inhibitionof STAT1 and NF-κB pathway, which contributes to theameliorated necroptosis of OLs. Taken together, ourstudy demonstrated that quercetin alleviated necroptosisof OLs in part by suppressing M1 macrophages/micro-glia polarization through inhibition of STAT1 and NF-κB pathway after SCI.

MethodsExperimental designMale Sprague-Dawley rats (200–250 g) were purchasedfrom the Laboratory Animal Center of Xi’an JiaotongUniversity. This study was carried out in accordancewith the recommendations of ‘Animal Care and Use,Committee of Xi’an Jiaotong University.’ The protocolwas approved by the ‘Committee of Xi’an Jiaotong Uni-versity.’ A total of 120 rats were equally randomized intothree groups: Sham group, SCI + saline group, and SCI+ quercetin group. The sham group received only lamin-ectomy, and the vehicle (1 ml sterile saline plus 1%DMSO) was administrated via intraperitoneal injection(i.p.) after SCI as SCI + saline group. According to theprevious finding that functional recovery correlatestreatment duration of quercetin after SCI [19], 7.5 mg/kg quercetin (Sigma-Aldrich, St. Louis, MO, USA) in1 ml vehicle was administered i.p. twice daily for 10 daysafter SCI.

Animal model of spinal cord injuryRats were anesthetized with 1% sodium pentobarbital(60 mg/kg), and bilateral laminectomy of vertebrae T8was performed to expose the spinal cord. The spinalcord crush model was made by a mechanical device [20].Briefly, we used the forceps (53327T, 66 Vision-TechCo., Ltd., China) mounted onto the device to clamp thecord, with the forceps tips of 0.5-mm width when fullyclosed, and the duration of the injury remained for 20 s.After crush, manual bladder expression was performedonce a day, until reestablishment of micturition reflex.Depending on the following experiments, rats weresacrificed at 10 and 21 days post injury (dpi).

Behavioral testsBasso-Beattie-Bresnahan scoresThe Basso-Beattie-Bresnahan (BBB) scores were used totest rat open-field locomotion at 1 day before, and 1, 3,7, 10, 14, and 21 days after spinal cord injury (SCI) fol-lowing methodology of BBB [21]. Functional scoresranges from 0 to 21 points in terms of the ankle, hip,knee, and trunk movement, as well as the coordinationof each rat. The locomotion scores was observed and re-corded by two independent observers blinded to theexperiments.

Rump-height Index assayThe rats were video-recorded during walking throughthe left to right side on a runway bar (160 cm long,10 cm wide, and 8 cm thick) at 1 day prior to and 1, 3,7, 10, 14, and 21 days post-SCI. The rump-height Index(RHI) is defined as the height of the rump, normalizedto the thickness of the beam measured along the samevertical line [22]. To minimize the variations of pre-

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 2 of 15

surgery RHI of each animal, the standardized RHI (div-iding post-injury value by pre-injury value) was appliedfor comparisons.

Tissue processingSeven days after SCI or at the end of the 21-day observa-tion period, rats were anesthetized and trans-cardiallyperfused with 200 ml of saline, followed by 400 ml of 4%PFA. Subsequently, 2-cm spinal cord segments encom-passing the lesion site were obtained and cryoprotectedin sucrose (30% in 0.1 M phosphate buffer, pH 7.4) untilthe blocks sank.Both serial transverse sections (12-μm-thick) and serial

sagittal sections (12-μm-thick) were cut on a cryostatand prepared for immunostaining.

ImmunohistochemistryAfter blocking in 0.01 M PBS containing 3% BSA and0.3% Triton X-100 for 30 min, sections were incubatedwith primary antibodies in humidified boxes overnightat 4 °C. The primary antibodies used in the experimentincluded rabbit anti-RIP3 (1:500, ENZO, New York, NY,USA), rat anti-MLKL (1:200, Millipore, Billerica, MA,USA), rabbit anti-phosphorylated MLKL (pMLKL) (1:300, Abcam, Cambridge, Cambridgeshire, UK), mouseanti-CC1 (1:600, Milipore, Billerica, MA, USA), rabbitanti-iNOS (1:200, Abcam, Cambridge, Cambridgeshire,UK), goat anti-Arginase1(1:200, Santa Cruz Biotechnol-ogy, Dallas, Texas, USA), mouse anti-glial fibrillaryacidic protein (GFAP,1:2000, Sigma-Aldrich, St. Louis,MO, USA), rabbit anti-MBP(1:200, Abcam, Cambridge,Cambridgeshire, UK), mouse anti-NF200 (1:200,Abcam,Cambridge, Cambridgeshire, UK), rabbit anti-CleavedCaspase-3(1:200, Cell Signaling, Danvers, MA, USA),goat anti-Iba-1(1:200,Abcam, Cambridge, Cambridge-shire, UK), mouse anti-PDGFRα (1:200, Cell Signaling,Danvers, MA, USA), rat anti-BrdU (1:300,Abcam, Cam-bridge, Cambridgeshire, UK), and rabbit anti-Olig2 (1:200, Abcam, Cambridge, Cambridgeshire, UK). The sec-tions were then incubated with the corresponding sec-ondary antibodies conjugated with Alexa Fluor 594(donkey anti-rabbit IgG or donkey anti-mouse IgG, 1:1000, Jackson ImmunoResearch, West Grove, PA, USA),Alexa Fluor 488 (donkey anti-rat IgG or donkey anti-mouse IgG, 1:1000, Jackson mmunoResearch, WestGrove, PA, USA) in a dark environment for at RT 1 h.The sections were counterstained with Hoechst 33342 toidentify all nuclei. Sections were examined and photo-graphed under a confocal microscope (LSM 800, Zeiss,Oberkochen, Germany).

BrdU incorporation and detectionFor detection of regeneration of OLs, BrdU (100 mg/kg,Sigma, St. Louis, MO, USA) injection was initiated at

24 h after injury, once a day for 9 days. Rats were sacri-ficed at 10 days after injury. HCl treatment (2 N HCl,30 min at 37 °C) was performed for BrdU staining.

Luxol fast blue stainingFor Luxol fast blue (LFB) staining, serial transverse cryo-sections (12 μm thickness) were stained as previouslydescribed. Briefly, slides were incubated in 0.1% LFB(Sigma, St. Louis, MO, USA) in acidified 95% ethanolovernight at 60 °C. The slides were then differentiatedand counterstained with 0.05% lithium carbonate andcresyl violet solution. LFB-stained sections at 2000 μmrostral and caudal to the injury site, as well as at the le-sion epicenter, were analyzed for myelin sparing. Thespared myelin in white matter was quantified by Imagepro-plus (Media Cybernetics, Silver Spring, USA).

Electron microscopic studyRats were deeply anesthetized with 1% sodium pentobar-bital intraperitoneally (60 mg/kg body weight) and per-fused with 200 ml saline, followed by 500 ml 4% PFAcontaining 0.05% glutaraldehyde. Then the injured cordwas removed and postfixed in the same fixative solutionfor 4 h at 4 °C. Serial coronal sections of 50 μm thick-ness were prepared with a vibratome (VT 1000S, Leica,Wetzlar, Hesse, Germany), then placed in 0.01 M PBScontaining 25% sucrose and 10% glycerol for 2 h forcryoprotection. The sections were then immersed in0.5% osmium tetroxide for 1 h, dehydrated with gradedethanol series, then in propylene oxide, and finally flat-embedded in Epon812 (SPI-CHEM). After trimmingunder a stereomicroscope, the sections were glued ontoblank resin stubs. Serial ultrathin sections were cut withan ultramicrotome (Leica EM UC6, Leica, Wetzlar,Hesse, Germany) and mounted on mesh grids. Theywere then counterstained with uranyl acetate and leadcitrate, observed, and photographed under an electronmicroscope (JEM-1230, JEOL Ltd., Akishima, Tokyo,Japan) equipped with CCD camera and its applicationsoftware (832 SC1000, Gatan, Pleasanton, CA, USA).

Axonal diameter and G ratio determinationAxonal diameter was determined from counting of about200 axons in white matter of the injured segments ofthree rats at 21 dpi using Image J software. G-ratio wasdetermined as the ratio of inner axonal diameter to thetotal outer diameter of each axon fiber of about thesame 200 axons of three rats at 21 dpi.

Primary microglial culture and polarizationMicroglia were isolated from P2 Sprague-Dawley ratpups as described. Briefly, the spinal cord was dissected,mechanically dissociated, and then digested with 0.125%trypsin/0.02% EDTA. Cells were plated onto poly-D-

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 3 of 15

lysine-coated culture flasks and cultured in DMEM(Gibco, Waltham, USA) containing 10% FCS at 37 °Cwith 5% CO2. The medium was exchanged every 3 daysand at about 8th day when cells became confluent, theflasks were shaken with a shaker at 260 rpm for 0.5 h toobtain microglia. The quality of microglia was analyzedby immunostaining of Iba1 (> 90% Iba1+).After about 7 days culture, microglial cells were

treated with LPS (100 ng/ml, Sigma, Santa Clara, USA)and IFN-γ (20 ng/ml, Peprotech, Rocky Hill, USA) forM1 polarization, or quercetin plus LPS and IFN-γ,followed by medium replacement. Twenty-four hourslater, the conditioned medium (M1 CM and Q-M1 CM)was obtained.

Primary oligodendrocytes cultureAfter microglia was harvested, with another overnightshaking at 280 rpm, cell suspension was harvested andseeded on tissue culture dishes (Corning, NY, USA) for40 min. The dishes were then gently swirled every 6 minto remove contaminating astrocytes and microglia. Thenon-adherent cells were seeded onto poly-D-lysine-coated culture dishes in DMEM containing 10% FCS atdensity of 5 × 105 cells/cm2 for 3 h, and the medium wasthen changed by oligodendrocyte precursor cells (OPCs)proliferation medium (chemically defined medium with10 ng/ml PDGF-AA and 10 ng/ml bFGF). Three dayslater, OPCs were isolated with DMEM/F12 containing0.01% EDTA, 0.2 mg/ml DNase I, and 5 mg/ml insulin,and seeded in differentiation medium (chemically de-fined medium with 30 nM triiodothyronine and 5 mg/ml N-acetyl-L-cysteine). Mature oligodendrocytes (OLs)were obtained 5 days later. Chemically defined mediumcontains high glucose DMEM/F12 supplemented with4 mM L-glutamine, 50 mg/ml transferrin, 1 mM sodiumpyruvate, 5 mg/ml insulin, and 10 ng/ml D-biotin. Themature OLs were identified by MBP staining.

Western blottingAbout 1.5 cm length of injured cord segments or pri-mary cultured microglia were harvested and homoge-nized with ice-cold radio-immunoprecipitation assay(RIPA) lysis buffer. Proteins were separated in 10%SDS–PAGE gel, and transferred onto polyvinylidenedifluoride (PVDF) membrane. The membranes were firstblocked with 5% non-fat milk in TBST for 1 h at RT,and incubated with primary antibodies for RIP3 (1:800,ENZO, New York, NY, USA), MLKL (1:500, Millipore,Billerica, MA, USA), pMLKL (1:500, Abcam, Cambridge,Cambridgeshire, UK), iNOS(1:500, Abcam, Cambridge,Cambridgeshire, UK), p-STAT1 (1:1000, Cell Signaling,Danvers, MA, USA), nuclear factor-kappa B (NF-κB)(P65,1:1000, Cell Signaling, Danvers, MA, USA),phospho-NF-κB p65/p-P65 (1:1000, Cell Signaling,

Danvers, MA, USA), or beta-actin (1:5000, Sigma, SantaClara, USA) at 4 °C overnight. The membranes were thenincubated with corresponding horseradish peroxidase-conjugated secondary antibodies (1:5000, Jackson Immu-noResearch, West Grove, PA, USA) at RT for 1 h. The im-munoreactive bands were scanned with the Bio-RadImage Lab system. The expression level was measured byImageJ software (NIH, USA) corresponding to the bandintensities.

Quantitative reverse transcription polymerase chainreactionTotal RNA was extracted from 1.5 cm length of injuredcord segments or microglia using TRizol (Invitrogen,Carlsbad, CA, USA) following the manufacturer’s in-structions. Complementary DNA synthesis was per-formed using a Reverse Transcriptase kit (Invitrogen,Carlsbad, CA, USA). Real-time PCR for TNFα, iNOS,CD86, CD206, Arginase1, IL-4, IL-10, IL-12, IL-1β, andTGFβ were performed with SYBR Green (TaKaRa, Da-lian, China) using Bio-Rad 96 Real-time PCR system(Bio-Rad Laboratories, Hercules, CA, USA), withGAPDH used as a reference control. The primer se-quences are listed in Table 1.

Propidium iodide stainingAfter the cells were treated with M1 CM and Q-M1CM, propidium iodide (PI) (5 μg/ml, Sigma, Santa Clara,USA) and Hoechst 33342 (3 μg/ml, Sigma, Santa Clara,USA) were added into the culture medium and incu-bated for 20 min at 37 °C. After washing with 0.01 MPBS three times, cells were fixed with 4% PFA for10 min at RT. Cells were then imaged and countedunder inverted fluorescence microscope (IX71, Olympus,Tokyo, Japan).

Table 1 PCR primer sequences

Gene Forward primer(5′-3′) Reverse primer (5′-3′)

TNFα ATGGGCTCCCTCTCATCAGT GCTTGGTGGTTTGCTACGAC

iNOS TCCTCAGGCTTGGGTCTTGT ATCCTGTGTTGTTGGGCTGG

CD86 AGACATGTGTAACCTGCACCAT TACGAGCTCACTCGGGCTTA

CD206 TCAACTCTTGGACTCACGGC CATGATCTGCGACTCCGACA

Arg1 CCAGTATTCACCCCGGCTAC GTCCTGAAAGTAGCCCTGTCT

IL-4 GTACCGGGAACGGTATCCAC TGGTGTTCCTTGTTGCCGTA

IL-10 CGACGCTGTCATCGATTTCTC CAGTAGATGCCGGGTGGTTC

IL-12 ATCATCAAACCGGACCCACC AGGAACGCACCTTTCTGGTT

IL-1β GCTTCCTTGTGCAAGTGTCT TCTGGACAGCCCAAGTCAAG

TGFβ CTGCTGACCCCCACTGATAC AGCCCTGTATTCCGTCTCCT

GAPDH AGTGCCAGCCTCGTCTCATA GGTAACCAGGCGTCCGATAC

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 4 of 15

ROS measurementDCFH-DA probe (Sigma, St. Louis, MO, USA) was usedto detect reactive oxygen species (ROS) levels. DCFH-DA (10 μM) was added to 96-well plates seeded withOLs incubated for 30 min at 37 °C. ROS levels were thenexamined by a multimode microplate reader (TECAN,Infinite M200, Mannedorf, Switzerland).

ATP measurementATP levels were measured using the Cell Titer-Glo assaykit (Promega, Madison, USA). Equivalent Cell Titer-GloReagent was added into medium for 5 min to induce celllysis, followed by incubating for 15 min to stabilize lumi-nescent signal. Luminescence was detected by multimodemicroplate reader (TECAN, infinite M200, Mannedorf,Switzerland).

Quantification of cellsIn our study, the method of profile counts followed bycalibration with the empirical method was adopted forunbiased data [23, 24].The 12-μm thickness of the sec-tion ensures that the plane of central canal is cutthrough on every slide, making random selection of theslide possible. For cells quantification, all of the positivecells in the defined area in all sections of the randomlychosen slide were counted. Then calibration of the pro-file numbers was made by empirical method. Cell count-ing was performed by an observer who was blinded tothe experiment design.For quantitative analysis of density of MBP and

NF200, serial transverse cryosections (12-μm thickness)were collected every millimeter section rostral and cau-dal 2000 μm to the lesion site. The longitudinal sectionswere also performed to confirm the density of MBP. Thedata were presented as a percentage relative to that insham control (100%).

StatisticsQuantification was performed by researchers who wereblinded to the group allocation. No data were excludedfrom the analysis, except for five rats evaluated by BBBscores. In the present study, data are expressed as themean ± standard error of the mean (SEM). Comparisonsbetween two groups were evaluated by the Student’s ttest [25]. For multiple comparisons with the three orfour groups, we performed one-way analysis of variance(ANOVA), followed by Tukey post-hoc tests, except forBBB scoring and RHI assay which were further analyzedby the Bonferroni post hoc test as recommended[26, 27]. Differences were considered statistically sig-nificant if *p < 0.05, **p < 0.01, or ***p < 0.001.

ResultsQuercetin improved functional recovery after SCILocomotor functional recovery was evaluated using BBBscoring and RHI assay at 1 day before injury as well as 1,3, 7, 10, 14, and 21 days after SCI. Compared with SCI +saline control, the quercetin-treated rats showed signifi-cant higher BBB scores from 7 dpi (n = 6, *p <0.05,Fig. 1a), which was consistent with the previous results[16]. Similarly, the rump height shown by RHI was sig-nificantly elevated from 7 dpi within quercetin-treatedgroup (n = 6 rats/group, *p < 0.05, Fig. 1b).

Quercetin prevented necroptosis of OLs without affectingapoptosis and regeneration of OLs after SCIGiven that the survival of OLs is crucial for functionalrecovery after SCI [4, 28], we then focused on death and

Fig. 1 Quercetin improved the locomotion recovery after SCI. a BBBscores at 1, 3, 7, 10, 14, and 21 days post SCI. The BBB scores in quercetin-treated group were significantly higher than those in control group at 7,10, 14, and 21 days after injury. b SRHI values at 1, 3, 7, 10, 14, and 21 dayspost SCI. The rump height was significantly elevated in quercetin-treatedrats at 7, 10, 14, and 21 days after injury. All data are expressed as mean±SEM. Differences among groups were determined with one-way ANOVAfollowed by the Bonferroni post hoc test. N= 6 in each group, *p< 0.05

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 5 of 15

regeneration of OLs. Apart from apoptosis, RIP3/MLKL-mediated necroptosis is activated after SCI and our pre-vious study showed that CC1/MLKL double-positivecells appeared after SCI [9], indicating the necroptosis ofOLs. To investigate the effect of quercetin on necropto-sis of OLs after SCI, double staining of RIP3/CC1,MLKL/CC1, and pMLKL/CC1 were performed at 10 dayspost injury. The data of double staining showed that thenumber of RIP3/CC1, MLKL/CC1, and pMLKL/CC1was reduced by 59.05 ± 3.57%, 56.92 ± 3.59%, 43.3 ±1.07%, and 71.11 ± 3.82% respectively after quercetintreatment (n = 6, *p < 0.05, Fig. 2a–f). Moreover, quer-cetin improved OLs survival by about 80% shown byCC1-positive cells in our model (n = 6, *p < 0.05, Fig. 2g, h).None of the RIP3-, MLKL-, and pMLKL-positive cells wereobserved in sham-operated cords (data not shown). The

above results indicated that quercetin treatment signifi-cantly reduced the numbers of necroptosis of OLs afterSCI.Further, we examined that whether quercetin could im-

prove OL preservation only or enhance both OL preserva-tion and OL regeneration after SCI. BrdU administrationscheme was shown in Fig. 3a. By co-immunostaining ofPDGFRα/BrdU and CC1/Olig2, we found that quercetinhad no significant effect on OL regeneration after SCI(Fig. 3b, c). In addition, by co-immunostaining ofPDGFRα and RIP3, MLKL, or pMLKL, we found thatthere are none of the double-positive cells (Fig. 3d), indi-cating that none of the OPCs occurred necroptosis.Moreover, we showed that just a small proportion of

OLs(about 2.3%) died through apoptosis in the first5 days after SCI, and from 5 dpi to 10 dpi, more OLs

Fig. 2 Quercetin inhibited necropotosis of OLs after SCI. a, c, e Double-immunostaining of CC1 with RIP3, MLKL, or pMLKL in saline or quercetin-treated rats at 10 dpi. Representative images are from sections 100 μm to the lesion epicenter. Arrows indicate double-positive cells. Scale bar =30 μm. b, d, f Quantification of CC1/RIP3-, CC1/MLKL-, and CC1/pMLKL-positive cells. Notice the decrease of double-positive cells in quercetin-treated rats. *p < 0.05 compared to SCI + saline control. g, h Immunostaining and quantification of CC1-positive cells in sham, saline, or quercetin-treated rats at 10 dpi. Scale bar = 100 μm. Notice the increase of CC1-positive cells in quercetin-treated rats, compared to saline-treated rats. Alldata are expressed as mean ± SEM. Differences among groups were determined with unpaired two-tailed t test (b, d, f) or one-way ANOVAfollowed by Tukey post-hoc test (h). N = 6 in each group, *p < 0.05, ** p<0.01

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 6 of 15

were eliminated by necroptosis (about 13%) rather thanapoptosis (about 3.1%) (Additional file 1: Figure S1a–c),indicating that necroptosis was the main way of OLsloss in our model of SCI. In addition, no significanteffect of quercetin on apoptosis of OLs was found(Additional file 1: Figure S1d-e).

Quercetin reduced myelin and axonal loss after SCIGiven that death of OLs results in demyelination andsecondary axon damage [29, 30], we examined the effectof quercetin on loss of myelin and axons after SCI.Spinal cords at 21 days after injury were processed forLFB staining and immunostaining to assess the myelin

loss(Fig. 4a, b). In both dorsal and lateral funiculus ofthe white matter at 2000 μm rostral from the lesion andlongitudinal sections, myelin and axonal loss was exten-sive in the SCI + saline control, whereas quercetin treat-ment significantly attenuated MBP-positive myelin andNF200-positive axonal loss (n = 6; **p < 0.01, *p < 0.05,Fig. 4a–d). Demyelination was also observed in the sameregion of white matter by electron microscopic analysis,and the data showed that SCI induced decompaction ofmyelin sheaths with a decreased g-ratio, an increasedlarge-diameter axons, and a decreased axonal number,whereas quercetin treatment attenuated these effects(n =6; *p < 0.05, Fig. 4e–h). These data indicated that

Fig. 3 Quercetin had no effect on regeneration and apoptosis of OLs after SCI. a BrdU administration scheme. b, c Co-immunostaining ofPDGFRα/BrdU and CC1/Olig2, and quantification of double-positive cells. Note that no significant effect of quercetin on OL regeneration after SCI.d Double-immunostaining of PDGFRα and RIP3, MLKL, or pMLKL. Notice that none of the double-positive cells was found. Scale bars = 50 μm. Alldata are expressed as mean ± SEM. Differences among groups were determined with unpaired two-tailed t test (b, c). N = 6 in each group

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 7 of 15

quercetin treatment markedly reduced the extent of bothmyelin and axon loss after SCI.

Quercetin influenced the macrophages/microglia M1-M2polarization balance after SCIConsidering that the ameliorated cell necroptosis needsa permissive immuno-microenvironment and quercetin

has been used as an anti-inflammatory agent [10, 31], wethen explored the effects of quercetin on macrophages/microglia-mediated inflammation after SCI. The activa-tion of macrophages/microglia can be divided to pro-inflammatory, damaging M1 phenotype, and anti-inflammatory, regeneration-promoting M2 phenotypeafter SCI [32]. Because the anti-inflammatory effect of

Fig. 4 Quercetin reduced myelin and axonal loss after SCI. a Spinal cords at 21 dpi were processed for Luxol fast blue. b Double-staining of MBP/NF200 in transverse sections. Immunofluorescence image of MBP and NF200 in lateral and dorsal funiculus of cords in sham, saline, or quercetin-treated rats at 21 dpi after SCI. Transverse cryosections were selected 2000 μm rostral to the lesion site. Scale bar = 100 μm. c, d Quantification ofMBP and NF200 intensity. Note that quercetin treatment attenuated the reduction of MBP and NF200 in the white matter after injury. e–hElectron microscopic images and quantification of the mean axonal numbers, mean g-ratios, and mean axonal diameters in spinal cord whitematter at 21 dpi. Arrows indicate degenerated myelin exhibiting onion-like appearance whose myelin lamellae were disorganized and loosened.Note that the increased number of axonal numbers, mean g-ratios, and decreased mean axonal diameters in quercetin treated rats. Scale bar =2 μm. All data are expressed as mean ± SEM. Differences among groups were determined with one-way ANOVA followed by Tukey post-hoc test(c, d, f, h). N = 6 in each group, *p < 0.05; **p < 0.01

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 8 of 15

quercetin has been reported in SCI [17, 33], we specu-lated that quercetin could inhibit macrophages/microgliapolarization to M1 phenotype. To determine the effectof quercetin on M1 polarization, we first examined themRNA levels of TNFα, iNOS, and CD86, the markers ofM1 macrophages/microglia at 10 days after injury. Ourdata showed that mRNA of TNFα, iNOS, and CD86 wasdecreased by 50.17 ± 2.43%, 45.23 ± 2.06%, and 59.14 ±3.51% respectively in the quercetin group (n = 6 rats/group, *p < 0.05, Fig. 5a–c). Meanwhile, quantification inbilateral areas 200 μm rostral and dorsal to lesion center,as shown in Fig. 5f, demonstrated that quercetin de-creased the number of iNOS-expressing macrophages/microglia by 59.1 ± 2.26% (*p < 0.05, Fig. 5d, e).We further examined the effect of quercetin on M2

polarization. Interestingly, the results showed that

quercetin increased the mRNA levels of Arginase1,IL-4, and CD206 by 45.09 ± 2.14%, 65.89 ± 5.34%, and61.9 ± 3.75% respectively (n = 6 rats/group, *p < 0.05,Fig. 6a–c). Because M2 macrophages/microglia wasmainly located in the epicenter after SCI [34, 35], weanalyzed the numbers of M2 in the inner border ofGFAP-immunoreactive zone (Additional file 2). Theresult showed that the Arginase1-positive macro-phages/microglia was increased by 46.44 ± 4.79% (*p <0.05, Fig. 6d, e). None of the iNOS- or Arginase1-positive cells was observed in sham-operated cords(data not shown). Collectively, the above resultsshowed that quercetin could inhibit macrophages/microglia polarization to M1 phenotype, while pro-mote M2 polarization, which may forms an environ-ment allowing for survival of OLs.

Fig. 5 Quercetin reduced the numbers of M1-type macrophages/microglia in injured spinal cord. a–c Real-time reverse-transcriptase polymerasechain reaction of M1-associated mRNA transcripts of TNFα, iNOS, and CD86 in saline or quercetin-treated rats at 10 dpi. d Double-staining ofiNOS and Iba-1 in quercetin-treated or saline-treated rats at 10 dpi. Representative images are from sections 20 μm to the lesion epicenter. Scalebar = 30 μm. e Quantification of iNOS-positive microglia/macrophages. Notice the decrease of iNOS-positive microglia/macrophages in quercetin-treated rats. *p < 0.05 compared to SCI + saline control. f The bilateral areas 200 μm rostral and caudal to the lesion site and the lesion epicenteron which quantification was performed. Scale bar = 80 μm. All data are expressed as mean ± SEM. Differences among groups were determinedwith unpaired two-tailed t test (a–c, e). N = 6 in each group, *p < 0.05

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 9 of 15

The mechanism of effect of quercetin on M1macrophages/microglia-induced necroptosis of OLsThe pro-inflammatory response is always accompaniedwith cell death after SCI [36]. To explore the direct evi-dence that M1 macrophages/microglia contributes tonecroptosis of OLs, we further cultured OLs and microgliaand induced necroptosis. MBP staining was performed toidentify the purity of OLs, and about 80% of the cells areMBP positive (data not shown). Quercetin-modified M1CM (Q-M1 CM) was obtained by addition of quercetin tomicroglia when polarizing them to M1 phenotype. Aftertreatment with conditioned medium of M1 microglia (M1CM), the ROS level of OLs increased 3.3-fold, ATP leveldecreased 2.4-fold, PI labeled cells increased 6.7-fold, whichwere the indicators of necroptosis [37, 38] (n = 3, **p < 0.01,Fig. 7a, b, g, h). The expression of RIP3, MLKL, and p-MLKL was also significantly enhanced by M1 CM (n = 3,**p < 0.01, *p < 0.05, Fig. 7c–f). The effects were significantly

blocked by Q-M1 CM, and quercetin alone has nosignificant effects on necroptosis of OLs (n = 3, *p < 0.05,Fig. 7a–h). These data suggested that quercetin inhibitednecroptosis of OLs induced by M1 macrophages/microglia.To determine whether this reduction resulted from

the suppressed polarization of M1 macrophages/micro-glia, mRNA levels of M1-related markers and M2-related markers were detected. The results showed thatquercetin decreased the mRNA levels of TNFα, IL-12,and IL-1β in M1 microglia, while increased mRNA levelsof IL-4, IL-10, and TGF-β (n = 3, ***p < 0.001, **p < 0.01,*p < 0.05, Fig. 8a, b). The data provides the direct evi-dence that quercetin prevented macrophages/microgliapolarized to M1 phenotype.Further, we explored the possible mechanism of the

above effect by examining the STAT1 and NF-κB signal-ing pathway, which was reported to be involved in tran-scriptional regulation of M1polarization of macrophages/

Fig. 6 Quercetin increased the numbers of M2-type macrophages/microglia after SCI. a–c Expression of M2-associated mRNA transcripts ofArginase1, IL-4, CD206 in saline, or quercetin-treated rats at 10 dpi. d Double-staining of Arginase1 and Iba-1 in quercetin-treated or saline-treatedrats at 10 dpi. Scale bar = 30 μm. e Quantification of the numbers of Arginase1-positive microglia/macrophages. Sections were taken from thelesion epicenter. All data are expressed as mean ± SEM. Differences among groups were determined with unpaired two-tailed t test (a–c, e).N = 6 in each group, *p < 0.05

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 10 of 15

microglia. In vitro, the expression of iNOS, pSTAT1, NF-κB, and p-NF-κB was decreased by 56.94 ± 2.49%, 58.31 ±5.86%, 56.6 ± 3.46%, and 53.33 ± 9.58% respectively whenquercetin was added to M1 microglia (n = 3, **p < 0.01,*p < 0.05, Fig. 8c, d). In vivo, SCI dramatically inducedthe expression of iNOS, pSTAT1, NF-κB, and p-NF-κB, and quercetin treatment significantly reduced

their expressions by 42.86 ± 4.01%, 51.28 ± 2.41%,30.43 ± 3.46%, and 57.37 ± 6.16% respectively (n = 3,**p < 0.01, *p < 0.05, Fig. 8e, f).

DiscussionIn this study, we demonstrated that quercetin treatmentimproved functional recovery after SCI. Quercetin also

Fig. 7 Quercetin inhibited M1 macrophages/microglia induced-necroptosis of OLs. a, b Quantification of intracellular ROS and ATP levels in OLs,or cells treated by quercetin alone, conditioned medium (CM) from M1 microglia (M1 CM) and quercetin-treated M1 microglia (Q-M1 CM). Noticethat quercetin significantly attenuated the effects of M1 CM on necroptosis of OLs. c–f Quantification of the expression levels of RIP3, MLKL, andpMLKL in control cells, quercetin alone-treated, M1 CM-treated, or Q-M1 CM-treated cells. β-actin was used as a loading control. Notice thatquercetin significantly decreased the expression of necroptotic markers induced by M1 CM. g, h PI staining and quantification of PI-positive cellsin control cells, or cells treated by quercetin alone, M1 CM, and Q-M1 CM. Notice that quercetin significantly decreased the number of PI-labeledcells induced by M1 CM. Scale bar = 50 μm. Results are mean ± SEM of three independent experiments. Differences among groups weredetermined with one-way ANOVA followed by Tukey post-hoc test (a, b, d–f, h). *p < 0.05; **p < 0.01

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 11 of 15

attenuated RIP3/MLKL-mediated necroptosis of OLsafter injury, while suppressed myelin and axon loss inthe white matter after SCI. We then explored the reasonand mechanism of increased number of rescued OLsafter SCI, and we found that quercetin alleviated macro-phages/microglia polarized to M1 phenotype, while

promoted M2 polarization after SCI. The new balance ofM1-M2 polarization of macrophages/microglia mayform a permissive environment allowing for OL survival.Our in vitro study provided direct evidence that quer-cetin prevented necroptosis of OLs induced by M1 mac-rophages/microglia.

Fig. 8 Effects of quercetin on M1 polarization of microglia and expression of pSTAT1, NF-κB, and pNF-κB. a, b The effect of quercetin on mRNAlevels of M1 related TNFα, IL-12, IL-1β, and M2-related IL-4, IL-10, and TGF-β in M1 microglia. Note that quercetin significantly decreased mRNAlevels of TNFα, IL-12, and IL-1β, while increased mRNA levels of IL-4, IL-10, and TGF-β. Data were expressed as mean ± SEM of three independentexperiments. c Expression of iNOS, pSTAT1, NF-κB, and pNF-κB in microglia treated by LPS + IFN-γ in the presence or absence of quercetin. dQuantification of protein levels. β-actin was used as a loading control. Data were expressed as mean ± SEM of three independent experiments. eExpression of iNOS, pSTAT1, NF-κB, and pNF-κB in sham, saline, or quercetin-treated rats at 10 dpi after SCI. f Quantification of protein levels. Datawere expressed as mean ± SEM of 6 rats. Differences among groups were determined with unpaired two-tailed t test (a, b, d) or one-way ANOVAfollowed by Tukey post-hoc test (f). *p < 0.05; **p < 0.01; ***p < 0.001

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 12 of 15

As numerous intact demyelinated axons are observedafter SCI [39], there is an urgent need to rescue myelinsheath. It was reported that p75NTR-mediated apoptosisof OLs can be induced by proNGF produced in activatedmacrophages/microglia after SCI [40, 41]. However, thisacute apoptosis of OLs could not explain the chronic,delayed demyelination after SCI [42], suggesting thatother types of OLs death accounted for the delayed de-myelination. In our previous studies, we demonstratedthat necroptosis is a chronic process and necroptosiscan occur in OLs after SCI [9, 10]; we thus focused onthe effects of quercetin on necroptosis of OLs in thepresent study. We found that quercetin inhibitednecroptosis of OLs and reduced myelin and axonal loss.Further, we examined that whether quercetin could im-prove OL preservation only or enhance both OL preser-vation and OL regeneration after SCI. In the previousstudy, quercetin has been found to promote proliferationand differentiation of OPCs after oxygen/glucosedeprivation-induced injury in vitro [43], which is a directeffect of quercetin on OPCs. However, in this study, nosignificant difference of OL regeneration between con-trol and quercetin-treated rats was found, and we specu-lated that the microenvironment is one of the possiblereasons for the different results between in vitro andin vivo. Moreover, no significant effect of quercetin onapoptosis of OLs was observed. Thus, in the presentstudy, we concluded that quercetin inhibited necroptosisof OLs without influencing regeneration and apoptosisof OLs.Spinal cord injury can activate quiescent microglia po-

larized to M1/M2 phenotypes as well as the recruitedmacrophages. The distinct phenotype may exert their re-spective roles in pathological event of SCI. The inflam-matory response in SCI is characterized by predominantand prolonged M1 macrophages/microglia polarization[35], which forms a detrimental environment for OLsurvival. Quercetin, a compound acting as anti-oxidativeand anti-inflammatory, has been shown to have neuro-protective effects partially by inhibiting inflammatory re-sponse after SCI [16]. In our study, we mainly focusedon the effect of quercetin on polarization of macro-phage/microglia and the data revealed that quercetinsuppressed macrophages/microglia polarization to M1phenotype, while promoted M2 polarization without af-fecting the total amount of macrophage/microglia.After hemorrhagic brain injury, reduced M1polarization

of microglia was reported by inhibiting TLR4 [44], and wespeculated that the difference between our study and theprevious research is that we combined LPS and IFNγ totreat cells. It is also known that activation of STAT1 andNF-κB signaling pathways can skew macrophages/micro-glia toward the M1 phenotype [45]. In this study, wefound that quercetin suppressed macrophages/microglia

polarized to M1 phenotype through inhibition of STAT1and NF-κB pathway, which was consistent with the previ-ous results from BV2 cell lines [46]. Nevertheless, the spe-cific mechanism for inhibition of M1 polarization byquercetin requires further study.Time window for administration of agents is pivotal to

therapeutic effect [47]. Although the previous studyshowed that neuroprotection can be found when quer-cetin was administered twice daily for a period of either3 or 10 days [19], we selected the latter because the M1macrophages/microglia-mediated inflammatory responseis a chronic and long-lasting period [35]. In addition, inour preliminary experiment, quercetin alone was admin-istrated to normal rats without SCI, and no effect wasfound.Since the functions of microglia and macrophages are

similar but not exactly the same [48, 49], a major chal-lenge after SCI is to distinguish microglia and recruitedmacrophages to determine their phenotype and function.In the present study, we investigated the effects of quer-cetin on polarization of microglia/macrophages, and onegeneral limitation of our research is that these two dis-tinct cells had not been distinguished. It was reportedthat application of genetic reporter mice and flow cy-tometry are the commonly used method to distinguishhost microglia and recruited macrophages after injury[50–52]. In the present study, we applied Sprague-Dawley rats instead of genetic mice. Data of cytometryanalysis from previous studies showed that CD11b+,CD45low, Ly6G−, and Ly6C− represent the vast majority(> 95%) of microglia after SCI [52, 53]. Therefore, wetried our best to differentiate the host microglia and re-cruited macrophages by flow cytometry, but we failed toget data due to our technical restrictions. However, thisimportant issue is worthy to be studied in future.

ConclusionThis study demonstrated that quercetin treatment allevi-ated necroptosis of OLs at least in part by inhibiting M1macrophages/microglia polarization after SCI. Its benefi-cial effects can be potentially useful as a therapeuticagent for clinical SCI.

Supplementary informationSupplementary information accompanies this paper at https://doi.org/10.1186/s12974-019-1613-2.

Additional file 1: Figure S1. The proportion of apoptotic andnecroptotic OLs and the effect of quercetin on apoptosis of OLs. a Theco-staining of Cleaved Caspase-3/MLKL/CC1 at different time points afterSCI. Scale bar = 50 μm. b-c Quantification of the proportion of apoptoticand necroptotic OLs at different time points after SCI. d-e Quantificationand immunostaining of Cleaved Caspase-3 and CC1 in quercetin treatedor SCI + saline control rats at 10 dpi. Note that no significant effect ofquercetin on apoptosis of OLs. Scale bar = 30 μm. All data are expressed

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 13 of 15

as mean ± SEM. Differences among groups were determined with un-paired two-tailed t test (e). N = 6 in each group.

Additional file 2: Figure S2. The area that the quantification ofArginase1-positive macrophages/microglia was performed.

AbbreviationsBBB: Basso-Beattie-Bresnahan; GFAP: Glial fibrillary acidic protein;IL: Interleukin; LFB: Luxol fast blue; MLKL: Mixed lineage kinase domain-likeprotein; OLs: Oligodendrocytes; OPCs: Oligodendrocyte precursor cells;PI: Propidium iodide; PVDF: Polyvinylidene difluoride; qRT-PCR: Quantitativereverse transcription polymerase chain reaction; RHI: Rump-height Index;RIP3: Receptor-interacting serine-threonine kinase 3; RIPA: Radio-immunoprecipitation assay; ROS: Reactive oxygen species; SCI: Spinal cordinjury; TLR: Toll-like receptor

AcknowledgmentsWe thank the support from Translational Medicine Center and Department ofSpine Surgery, Hong Hui Hospital, Xi’an Jiaotong University. All experimentswere conducted in compliance with the ARRIVE guidelines.

Authors’ contributionsHF and HBT wrote the paper and performed the most of experiments. LQS,SCL, DGH, XC, ZC, MY, and XHY contributed morphological study andanalyzed the data. HY and DJH designed the research and revised themanuscript. All authors read and approved the final manuscript.

FundingThis work was supported by grants from the National Natural ScienceFoundation of China (NSFC, Grant code: 81701204) and Project funded byChina Postdoctoral Science Foundation (Grant code: 2018 M643703) to HongFan, NSFC (Grant code: 81830077, 81472098) to Ding-Jun Hao, NSFC (Grantcode: 81571208) to Hao Yang, NSFC (Grant code: 81802690) to Shi-ChangLiu, and Research Project of Xi’an Health Commission (Grant code:J201902019) to Hai-Bin Tang.

Availability of data and materialsAll data generated or analyzed in this study are included in this publishedarticle and its additional files.

Ethics approval and consent to participateAll protocols were approved by the Animal Care and Use Committee ofXi’an Jiaotong University, and conformed to the Guide for the Care and Useof Laboratory Animals by the National Institutes of Health.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Author details1Shaanxi Spine Medicine Research Center, Translational Medicine Center,Department of Spine Surgery, Hong Hui Hospital, Xi’an Jiaotong University,555 You Yi Dong Road, Xi’an 710054, Shaanxi, China. 2Institute ofNeurosciences, Fourth Military Medical University, Xi’an 710032, Shaanxi,China. 3Department of Laboratory Medicine, Xi’an Central Hospital, Xi’anJiaotong University, 161 Xi Wu Road, Xi’an 710003, Shaanxi, China.4Department of Bone Microsurgery, Hong Hui Hospital, Xi’an JiaotongUniversity, 555 You Yi Dong Road, Xi’an 710054, Shaanxi, China.

Received: 3 April 2019 Accepted: 9 October 2019

References1. Rong H, Zhao Z, Feng J, Lei Y, Wu H, Sun R, Zhang Z, Hou B, Zhang W, Sun

Y, et al. The effects of dexmedetomidine pretreatment on the pro- and anti-inflammation systems after spinal cord injury in rats. Brain Behav Immun.2017;64:195–207.

2. Profyris C, Cheema SS, Zang D, Azari MF, Boyle K, Petratos S. Degenerativeand regenerative mechanisms governing spinal cord injury. Neurobiol Dis.2004;15:415–36.

3. Beattie MS, Hermann GE, Rogers RC, Bresnahan JC. Cell death in models ofspinal cord injury. Prog Brain Res. 2002;137:37–47.

4. Papastefanaki F, Matsas R. From demyelination to remyelination: the roadtoward therapies for spinal cord injury. Glia. 2015;63:1101–25.

5. Beattie MS, Li Q, Bresnahan JC. Cell death and plasticity after experimentalspinal cord injury. Prog Brain Res. 2000;128:9–21.

6. Rathnasamy G, Ling EA, Kaur C. Iron and iron regulatory proteins inamoeboid microglial cells are linked to oligodendrocyte death in hypoxicneonatal rat periventricular white matter through production ofproinflammatory cytokines and reactive oxygen/nitrogen species. JNeurosci. 2011;31:17982–95.

7. Park E, Velumian AA, Fehlings MG. The role of excitotoxicity in secondarymechanisms of spinal cord injury: a review with an emphasis on theimplications for white matter degeneration. J Neurotrauma. 2004;21:754–74.

8. Ito Y, Ofengeim D, Najafov A, Das S, Saberi S, Li Y, Hitomi J, Zhu H, Chen H,Mayo L, et al. RIPK1 mediates axonal degeneration by promotinginflammation and necroptosis in ALS. Science. 2016;353:603–8.

9. Fan H, Tang HB, Kang J, Shan L, Song H, Zhu K, Wang J, Ju G, Wang YZ.Involvement of endoplasmic reticulum stress in the necroptosis of microglia/macrophages after spinal cord injury. Neuroscience. 2015;311:362–73.

10. Fan H, Zhang K, Shan L, Kuang F, Chen K, Zhu K, Ma H, Ju G, Wang YZ.Reactive astrocytes undergo M1 microglia/macrohpages-inducednecroptosis in spinal cord injury. Mol Neurodegener. 2016;11:14.

11. Kroner A, Greenhalgh AD, Zarruk JG, Passos Dos Santos R, Gaestel M, DavidS. TNF and increased intracellular iron alter macrophage polarization to adetrimental M1 phenotype in the injured spinal cord. Neuron. 2014;83:1098–116.

12. Boots AW, Haenen GR, Bast A. Health effects of quercetin: from antioxidantto nutraceutical. Eur J Pharmacol. 2008;585:325–37.

13. Abarikwu SO, Pant AB, Farombi EO. Dietary antioxidant, quercetin, protectssertoli-germ cell coculture from atrazine-induced oxidative damage. JBiochem Mol Toxicol. 2012;26:477–85.

14. Hogan S, Canning C, Sun S, Sun X, Zhou K. Effects of grape pomaceantioxidant extract on oxidative stress and inflammation in diet inducedobese mice. J Agric Food Chem. 2010;58:11250–6.

15. Wang Y, Li W, Wang M, Lin C, Li G, Zhou X, Luo J, Jin D. Quercetin reducesneural tissue damage and promotes astrocyte activation after spinal cordinjury in rats. J Cell Biochem. 2018;119:2298–306.

16. Jiang W, Huang Y, Han N, He F, Li M, Bian Z, Liu J, Sun T, Zhu L. Quercetinsuppresses NLRP3 inflammasome activation and attenuates histopathologyin a rat model of spinal cord injury. Spinal Cord. 2016;54:592–6.

17. Zhang Y, Yi B, Ma J, Zhang L, Zhang H, Yang Y, Dai Y. Quercetin promotesneuronal and behavioral recovery by suppressing inflammatory responseand apoptosis in a rat model of intracerebral hemorrhage. Neurochem Res.2015;40:195–203.

18. Schultke E, Kendall E, Kamencic H, Ghong Z, Griebel RW, Juurlink BH.Quercetin promotes functional recovery following acute spinal cord injury. JNeurotrauma. 2003;20:583–91.

19. Schultke E, Kamencic H, Skihar VM, Griebel R, Juurlink B. Quercetin in ananimal model of spinal cord compression injury: correlation of treatmentduration with recovery of motor function. Spinal Cord. 2010;48:112–7.

20. Fan H, Chen K, Duan L, Wang YZ, Ju G. Beneficial effects of early hemostasison spinal cord injury in the rat. Spinal Cord. 2016;54:1058.

21. Basso DM, Beattie MS, Bresnahan JC. A sensitive and reliable locomotorrating scale for open field testing in rats. J Neurotrauma. 1995;12:1–21.

22. Sedy J, Urdzikova L, Jendelova P, Sykova E. Methods for behavioral testingof spinal cord injured rats. Neurosci Biobehav Rev. 2008;32:550–80.

23. Coggeshall RE, Lekan HA. Methods for determining numbers of cells andsynapses: a case for more uniform standards of review. J Comp Neurol.1996;364:6–15.

24. Coggeshall RE, La Forte R, Klein CM. Calibration of methods for determiningnumbers of dorsal root ganglion cells. J Neurosci Methods. 1990;35:187–94.

25. Hazra A, Gogtay N. Biostatistics series module 3: comparing groups:numerical variables. Indian J Dermatol. 2016;61:251–60.

26. Sedgwick P. One way analysis of variance: post hoc testing: smoking inpregnancy. BMJ. 2014;349:g7067.

27. Scheff SW, Saucier DA, Cain ME. A statistical method for analyzing ratingscale data: the BBB locomotor score. J Neurotrauma. 2002;19:1251–60.

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 14 of 15

28. Gaudet AD, Fonken LK. Glial cells shape pathology and repair after spinalcord injury. Neurotherapeutics. 2018;15:554–77.

29. Wang X, Xu JM, Wang YP, Yang L, Li ZJ. Protective effects of BMP-7 againsttumor necrosis factor alpha-induced oligodendrocyte apoptosis. Int J DevNeurosci. 2016;53:10–7.

30. Pomeshchik Y, Kidin I, Korhonen P, Savchenko E, Jaronen M, Lehtonen S,Wojciechowski S, Kanninen K, Koistinaho J, Malm T. Interleukin-33 treatmentreduces secondary injury and improves functional recovery after contusionspinal cord injury. Brain Behav Immun. 2015;44:68–81.

31. Li Y, Yao J, Han C, Yang J, Chaudhry MT, Wang S, Liu H, Yin Y. Quercetin,Inflammation and Immunity. Nutrients. 2016;8:167.

32. Xu F, Huang J, He Z, Chen J, Tang X, Song Z, Guo Q, Huang C. Microglialpolarization dynamics in dorsal spinal cord in the early stages followingchronic sciatic nerve damage. Neurosci Lett. 2016;617:6–13.

33. Song Y, Liu J, Zhang F, Zhang J, Shi T, Zeng Z. Antioxidant effect ofquercetin against acute spinal cord injury in rats and its correlation with thep38MAPK/iNOS signaling pathway. Life Sci. 2013;92:1215–21.

34. Ma SF, Chen YJ, Zhang JX, Shen L, Wang R, Zhou JS, Hu JG, Lu HZ.Adoptive transfer of M2 macrophages promotes locomotor recovery inadult rats after spinal cord injury. Brain Behav Immun. 2015;45:157–70.

35. Gensel JC, Zhang B. Macrophage activation and its role in repair andpathology after spinal cord injury. Brain Res. 2015;1619:1–11.

36. Moon YJ, Lee JY, Oh MS, Pak YK, Park KS, Oh TH, Yune TY. Inhibition ofinflammation and oxidative stress by Angelica dahuricae radix extractdecreases apoptotic cell death and improves functional recovery after spinalcord injury. J Neurosci Res. 2012;90:243–56.

37. Degterev A, Huang Z, Boyce M, Li Y, Jagtap P, Mizushima N, Cuny GD,Mitchison TJ, Moskowitz MA, Yuan J. Chemical inhibitor of nonapoptoticcell death with therapeutic potential for ischemic brain injury. Nat ChemBiol. 2005;1:112–9.

38. Yang Z, Wang Y, Zhang Y, He X, Zhong CQ, Ni H, Chen X, Liang Y, Wu J,Zhao S, et al. RIP3 targets pyruvate dehydrogenase complex to increaseaerobic respiration in TNF-induced necroptosis. Nat Cell Biol. 2018;20:186–97.

39. McDonald JW, Belegu V. Demyelination and remyelination after spinal cordinjury. J Neurotrauma. 2006;23:345–59.

40. Yune TY, Lee JY, Jung GY, Kim SJ, Jiang MH, Kim YC, Oh YJ, Markelonis GJ,Oh TH. Minocycline alleviates death of oligodendrocytes by inhibiting pro-nerve growth factor production in microglia after spinal cord injury. JNeurosci. 2007;27:7751–61.

41. Lee JY, Kang SR, Yune TY. Fluoxetine prevents oligodendrocyte cell deathby inhibiting microglia activation after spinal cord injury. J Neurotrauma.2015;32:633–44.

42. Blight AR. Delayed demyelination and macrophage invasion: a candidate forsecondary cell damage in spinal cord injury. Cent Nerv Syst Trauma. 1985;2:299–315.

43. Wu X, Qu X, Zhang Q, Dong F, Yu H, Yan C, Qi D, Wang M, Liu X, Yao R.Quercetin promotes proliferation and differentiation of oligodendrocyteprecursor cells after oxygen/glucose deprivation-induced injury. Cell MolNeurobiol. 2014;34:463–71.

44. Lan X, Han X, Li Q, Li Q, Gao Y, Cheng T, Wan J, Zhu W, Wang J.Pinocembrin protects hemorrhagic brain primarily by inhibiting toll-likereceptor 4 and reducing M1 phenotype microglia. Brain Behav Immun.2017;61:326–39.

45. Wang N, Liang H, Zen K. Molecular mechanisms that influence themacrophage m1-m2 polarization balance. Front Immunol. 2014;5:614.

46. Chen JC, Ho FM, Pei-Dawn Lee C, Chen CP, Jeng KC, Hsu HB, Lee ST, WenTung W, Lin WW. Inhibition of iNOS gene expression by quercetin ismediated by the inhibition of IkappaB kinase, nuclear factor-kappa B andSTAT1, and depends on heme oxygenase-1 induction in mouse BV-2microglia. Eur J Pharmacol. 2005;521:9–20.

47. Young W. The therapeutic window for methylprednisolone treatment ofacute spinal cord injury: implications for cell injury mechanisms. Res PublAssoc Res Nerv Ment Dis. 1993;71:191–206.

48. Zhou X, He X, Ren Y. Function of microglia and macrophages in secondarydamage after spinal cord injury. Neural Regen Res. 2014;9:1787–95.

49. Davies CL, Miron VE. Distinct origins, gene expression and function ofmicroglia and monocyte-derived macrophages in CNS myelin injury andregeneration. Clin Immunol. 2018;189:57–62.

50. Saederup N, Cardona AE, Croft K, Mizutani M, Cotleur AC, Tsou CL,Ransohoff RM, Charo IF. Selective chemokine receptor usage by central

nervous system myeloid cells in CCR2-red fluorescent protein knock-inmice. PLoS One. 2010;5:e13693.

51. Fenrich KK, Weber P, Rougon G, Debarbieux F. Long- and short-termintravital imaging reveals differential spatiotemporal recruitment andfunction of myelomonocytic cells after spinal cord injury. J Physiol. 2013;591:4895–902.

52. Zarruk JG, Greenhalgh AD, David S. Microglia and macrophages differ intheir inflammatory profile after permanent brain ischemia. Exp Neurol. 2018;301:120–32.

53. Greenhalgh AD, David S. Differences in the phagocytic response ofmicroglia and peripheral macrophages after spinal cord injury and its effectson cell death. J Neurosci. 2014;34:6316–22.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Fan et al. Journal of Neuroinflammation (2019) 16:206 Page 15 of 15